Angus Dorbie (dorbie++at++bitch.reading.sgi.com)
Wed, 7 Aug 1996 09:13:15 +0100
I can get 11 Mtris/sec with no backface culling and 6.5 Mtris/sec with
backface culling.
With textured, lit, gouraud, shaded, zbuffered, I get 6.2 Mtris/sec
dropping to 4.4 Mtris/sec.
So the big question is, how much extra does 2 sided lighting cost, because
with the two options your backface culling saves nothing and costs a lot
before we even consider host or memory limits.
The answer, with textured, lit, gouraud, shaded, zbuffered & two sided
lighting I get 3 Mtris/sec.
So what does this mean?
Even dropping to 3 Mtris/sec your _still_ better off with two sided
lighting because your 4.4 Mtris/sec backfaced is effectively half that
figure because your application draws everything twice (you may save a
little due to culling but I doubt you'll save much)
Three final notes:
1) for the benchmark the tris here were display listed, doing anything else
would produce host limits which would favour the twosided approach even more.
2) the above tests were not fill limited, any fill limit would even things
out a lot.
4) Performance for operations like back face culling is being improved for
future software releases but maybe twosided lighting will also improve.
Rgds,
Angus
On Aug 6, 5:45pm, Chris Cederwall wrote:
> Subject: Re: 2 faces or 2-sided lighting ?
> -> I've already posted this question but I had no answer.
> ->
> -> Which of these 2 modelisation is the most performant ? (on RE2 and iR)
> -> - 2 faces with opposite normals,
> -> - 1 face with 2-sided lighting.
> ->
>
>
> People can correct me they feel otherwise..... but
>
> It is considered most functional to have 2 faces with single sided
> lighting and *backface* removal on.
>
> If your object is single sided, when you look at it from the back side
> it will not be drawn when backface removal is enabled. Enabling Backface
> removal will usually gain you 20% increase in drawing times on average.
>
>
> Cheers,
>
> --
>
> Chris Cederwall
> ceder++at++pbi.net
> 415.442-4952
>
> ------------------------------------------------------------------
> * the views expressed are only views and are not neccesarily statements
> that reflect pacific bell internet services or even complete sentences.
>
> =======================================================================
> List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
>-- End of excerpt from Chris Cederwall
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:53:19 PDT